EP0355158B1 - Procede et appareil de traitement de signaux de donnees echantillonnees - Google Patents

Procede et appareil de traitement de signaux de donnees echantillonnees Download PDF

Info

Publication number
EP0355158B1
EP0355158B1 EP89904671A EP89904671A EP0355158B1 EP 0355158 B1 EP0355158 B1 EP 0355158B1 EP 89904671 A EP89904671 A EP 89904671A EP 89904671 A EP89904671 A EP 89904671A EP 0355158 B1 EP0355158 B1 EP 0355158B1
Authority
EP
European Patent Office
Prior art keywords
vectors
preprocessed
samples
unprocessed
thru
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
EP89904671A
Other languages
German (de)
English (en)
Other versions
EP0355158A1 (fr
Inventor
Douglas W. Chabries
Richard W. Christiansen
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Publication of EP0355158A1 publication Critical patent/EP0355158A1/fr
Application granted granted Critical
Publication of EP0355158B1 publication Critical patent/EP0355158B1/fr
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications
    • G01S13/89—Radar or analogous systems specially adapted for specific applications for mapping or imaging
    • G01S13/90—Radar or analogous systems specially adapted for specific applications for mapping or imaging using synthetic aperture techniques, e.g. synthetic aperture radar [SAR] techniques
    • G01S13/9004—SAR image acquisition techniques
    • G01S13/9017—SAR image acquisition techniques with time domain processing of the SAR signals in azimuth

Definitions

  • the invention relates to an electronic data processing system for the addition of shifted preprocessed data from a memory.
  • the invention is particularly concerned with ground mapping methods and a system which is used in synthetic aperture radar.
  • an airplane flies over the ground that is to be mapped and transmits a sequence of radar pulses at the ground. Return radar signals from the transmitted pulses are sampled in the airplane; those samples are then compressed in the airplane via various data compression techniques; and the compressed data is then transmitted from the airplane to a receiving station on the ground. Alternatively, the original uncompressed samples are sent to the ground station. Data compression is performed to reduce the amount of information that needs to be transmitted, and thereby reduce the bandwidth of the transmission channel.
  • the compressed data is received in the ground station, it is then decompressed in order to reconstruct, to close approximation, the original data samples. Then, the reconstructed data samples, or the original samples if sent in uncompressed form, are further processed by fast Fourier transform methods (FFT methods) and inverse FFT methods in order to produce the ground map. Alternatively, the reconstructed data samples or original data samples are convolved with a sampled data function in order to produce the ground map.
  • FFT methods fast Fourier transform methods
  • inverse FFT methods inverse FFT methods
  • US-A-4,679,164 discloses a digital high speed programmable convolver using a circuit module which multiplies one or more variables with a single constant and adds the results together.
  • the electronic data processing system includes an electronic memory means which stores a set of preprocessed vectors V1*f(m,%) thru V n *f(m,%) where f(m,%) is a sampled data function having any number of dimensions m,..., * is a convolution operator, and V1 thru V N are a finite set of N unprocessed vectors each of which represents an anticipated group of input signal samples; a means for generating a sequence of samples from an input signal that is to be processed and for compressing said sequence of samples into a smaller sequence of index signals that correspond to the indexes 1 thru N of said unprocessed vectors; a means for receiving said smaller sequence of index signals and for reading from said memory means, those preprocessed vectors whose indexes match the received index signals; and a means for adding together the read preprocessed vectors while maintaining a predetermined offset between them as they are added.
  • visual images in a synthetic aperture radar system are formed by the following steps. Initially, before any radar signals are sent, a finite set of preprocessed vectors V1*f (m, n), V2*f(m, n)... V N *f(m,n) are stored in an electronic memory M.
  • V1, V2,...V N are respective unprocessed vectors, each of which represents an anticipated group of data samples from the radar return signal;
  • f(m,n) is a two-dimensional sampled data function with variations in each of two dimensions as represented by m and n; and * is a convolution operator.
  • a sequence of spaced apart radar pulses is transmitted from a flying object; and a return signal from the transmitted pulses is sampled.
  • Those return signal samples are ordered in an array; and that array is then compressed into a smaller array of indexes in which each index i identifies one of the unprocessed vectors V1, V2,...V N .
  • those preprocessed vectors whose indexes i match the indexes in the array of indexes are read from the memory M, and the read preprocessed vectors are added together while maintaining a predetermined offset between them as they are added.
  • Visual images are formed by sending the result of the offset additions to a CRT display.
  • FIG. 1 an overview of a preferred method for forming visual images via radar in accordance with the invention will be described. Additional details of each step in the FIG. 1 method, as well as apparatus for performing that method, will subsequently be described in conjunction with FIGs. 2-10.
  • a finite set of preprocessed vectors V1*f(m,n), V2*f(m,n),...V N *f(m,n) are stored in an electronic memory M.
  • This step is indicated by reference numerals 10a thru 10e.
  • the symbols V1, V2,...V N are respective unprocessed vectors, each of which represents a group of anticipated data samples
  • f(m,n) is a two-dimensional sampled data function with variations in each of two dimensions as represented by m and n
  • * is a convolution operator.
  • Forming and storing the preprocessed vectors V1*f (m,n), V2*f(m,n),...V N *f(m,n) is performed just one time. That step is done first in preparation for the remaining steps of the FIG. 1 process. Additional details on this forming and storing step 10a-10e, including a specific example, are described herein in conjunction with FIGs. 4, 6 and 7.
  • a sequence of spaced-apart radar pulses is transmitted from a flying object such as an airplane or a satellite. This is indicated by reference numeral 20a.
  • a return signal is motion compensated, frequency shifted, and sampled multiple times. Each such sample can be a real or complex number.
  • Those samples of the return signal are arranged in an array. All of this is indicated by reference numeral 20b. Samples of the return signal from one pulse make up one row of the array; samples of the return signal from the next pulse make up the next row of the array; etc.
  • each index i identifies one of the unprocessed vectors V1, V2,...V N .
  • This conversion step is called vector quantization.
  • a group of adjacent samples in the sample array is compared to each of the unprocessed vectors V1, V2,...V N . This is done to determine which of the vectors V1, V2,. ..V N most closely matches the group of samples from the sample array. Additional details of this step are described herein in conjunction with FIGs. 3, 4 and 5.
  • reference numerals 31a and 31b indicate two of the spaced-apart pulses that are transmitted; and reference numerals 32a and 32b indicate the corresponding samples that are taken of the return signals.
  • each transmitted pulse can be a modulated chirp of the form given by Eq. 1 in FIG. 2.
  • the carrier frequency f c is 9.35 GHz
  • the chirp length is 2.7 microseconds
  • the repetition rate is one chirp per 15 microseconds
  • the chirp rate is 33.33 megahertz per microsecond.
  • the return signal can be sampled at any predetermined rate, such as one sample every 10 nanoseconds.
  • FIG. 3 shows the details of how the samples of the return signal are arranged in an array. All of the samples 32a lie in the order in which they were taken along the first row of the array; all of the samples 32b lie in the order in which they were taken along the second row of the array; etc.
  • Each sample is a real or imaginary number, and it is called a pixel. It consists of a certain number of bits, such as eight, twelve or sixteen.
  • FIG. 3 shows the details of how the array of pixels in FIG. 3 is converted into a smaller array of indexes i.
  • This conversion step utilizes the unprocessed vectors V1, V2,...V N as shown in FIG. 4.
  • FIG. 5 shows the array of indexes which results from the conversion step.
  • each of the unprocessed vectors V1, V2,...V N consists of nine pixels which are arranged in a 3X3 array. And, all of the pixels in each unprocessed vector V1, V2,...V N are preset to certain respective values.
  • each of the pixels A thru I of vector V1 is shown as having a value of P1; each of the pixels A, B, C of vector V2 is shown as having a value of P2, while the remaining pixels have a value of P3; etc.
  • a 3X3 group of pixels G1 in the FIG. 3 array is compared to all of the unprocessed vectors V1, V2,...V N . This enables a determination to be made as to which of the vectors V1, V2,...V N most closely matches that sample group G1. Once that determination is made, the index of the matching vector is placed in the FIG. 5 array at a location which corresponds to the location of the sample group G1 in the FIG. 3 array. Then the comparison steps are repeated for another sample group.
  • indexes i1, i2, i3 respectively correspond to sample groups G1, G2, G3 in FIG. 3.
  • index i1 is shown as having a value 2; and that indicates that the unprocessed vector V2 most closely resembles the samples in group G1.
  • index i2 is shown as having a value of 5 which indicates that sample group G2 most closely resembles vector V5.
  • f(m,n) is assumed to be of the form f1(m)f2(n).
  • each row of vector V i is convolved with a sampled data function f1(m). This results in a new vector as shown in FIG. 6.
  • each column of the FIG. 6 vector is convolved with another sampled data function f2(n); and the result is shown in FIG. 7.
  • function f1(m) is the sampled impulse response of a range filter.
  • the transmitted pulse is the modulated chirp of the form given by Eq. 1 in FIG. 2
  • the corresponding range filter will have a sampled impulse response as given by Eq. 2 of FIG. 6.
  • function f2(n) is a sampled impulse response of an azimuth filter.
  • One such azimuth filter sampled impulse response is given by Eq. 3 in FIG. 7.
  • the sampled impulse response f2(n) for the azimuth filter actually varies as a function of the range index m. Consequently, the total number of preconvolved FIG. 7 vectors which could be stored equals the number of unprocessed vectors times the number of sample points in the range filter. Preferably, however, the total number of preconvolved FIG. 7 vectors which are actually stored is reduced by holding the azimuth filter function constant at its average value over several values of m. For example, in Eq. 3, m can be set equal to 25, 75, 125,... for all actual values of m between 1-50, 51-100, 101-150,... respectively.
  • the convolution step V i *f1(m) increases the number of columns in the unprocessed vector V i . If vector V i has X1 columns and function f1(m) has Y1 sample points, then the convolved vector will have X1+Y1-1 columns. Similarly, FIG. 7 shows that the convolution step V1*[f1(m)f2(n)] increases the number of rows in the FIG. 6 result. If V i *f1(m) has X2 rows and function f2(n) has Y2 sample points, then the convolved vector V i *[f1(m)f2(n)] will have X2+Y2-1 rows. Preferably, the number of sample points in each of the functions f1(m) and f2(n) is between ten and one thousand.
  • FIG. 8 it shows how the preprocessed vectors V i *f(m,n) which are read from memory M are added together in an offset fashion.
  • the symbols i1 and i2 indicate the indexes which occur in row 3 column 4 and row 3 column 5 respectively of the index array of FIG. 5.
  • symbol V i1 *f(m,n) indicates the preprocessed vector in memory M which results from performing the convolution steps of FIGs. 6 and 7 on the unprocessed vector V i1 .
  • symbol V i2 *f(m,n) indicates the preprocessed vector in memory M which results from performing the convolution steps of FIGs. 6 and 7 on the unprocessed vector V i2 .
  • One such offset addition occurs for each of the preprocessed vectors whose index is in the FIG. 5 index array.
  • these additions occur in a left-to-right sequence for all of the preprocessed vectors whose indexes are in a single row of the FIG. 5 array; then the additions continue for all the preprocessed vectors whose indexes are in the next succeeding row; etc.
  • those pixels in FIG. 8 which are indicated by reference numeral 60 will become completely processed; next, the pixels which are indicated by reference numeral 61 will become completely processed; etc.
  • the resulting visual image will grow in size in an orderly sequence of three rows at a time.
  • a primary feature of the above described method is that it significantly reduces the time and complexity in producing visual images when compared to the prior art. This improvement occurs because in the disclosed method, no FFT operation or convolution operation is performed in real time on the samples of the radar return signals or on any decompressed radar return samples. In fact, no data decompression occurs at all, and no multiplications occur at all. In the disclosed method, only data compression and offset additions occur in real time. Consequently, the return signal samples can readily be processed as fast as they are accumulated.
  • FIG. 9 the architecture of an electronic system for producing visual images in real time in accordance with the above method will be described. It includes a digital computer 71, a memory 72, another digital computer 73, another memory 74, an amplitude detector 75, a CRT display 76, and the memory M in which the preprocessed vectors V i *f(m,n) are stored. All of these modules are interconnected via data buses 76a-76g as shown.
  • samples of the return signal are sent on bus 76a to computer 71. That computer temporarily stores the samples in memory 72, and it quantizes them into an array of indexes as was described in conjunction with FIG. 5. Those indexes i1, i2,... are sent via bus 76c to computer 73. In response, computer 73 reads the corresponding preprocessed vectors V i1 *f(m,n), V i2 *f(m,n),... from memory M, and it adds the read vectors together in accordance with the offset addition of FIG. 8. Results of that addition are sent via bus 76e to memory 74; and from there, they are sent via bus 76f to the amplitude detector 75 and then to the CRT 75 for visual display.
  • FIG. 10 A flow chart of the program which computer 73 executes in carrying out its above described task is shown in FIG. 10. To begin, the program waits for a START signal on bus 76c which indicates that the receipt of a sequence of indexes is about to begin. Upon receipt of the START signal, a row counter register and a column counter register in computer 73 are initialized to one. This is indicated by reference numeral 81. Then, the computer 73 waits for the receipt of an index signal.
  • the program When an index signal i is received, the program reads the corresponding preprocessed vector V i *f(m,n) from memory M. This is indicated by reference numeral 82. Then the program performs an offset addition on the pixels of that vector as was described in conjunction with FIG. 8, and it stores the resulting sum in memory 74. This is indicated by reference numeral 83. Next, the program updates the row and column registers such that they indicate the location, in the FIG. 5 index array, of the next index in sequence. This is indicated by reference numeral 84. Thereafter, when that next index is received, all of the above steps are repeated.
  • vector quantization is used to compress the samples of the radar return signal.
  • any other data compression method may be used which results in a finite number of indexes that correspond to the indexes of the preconvolved vectors.
  • each of the unprocessed vectors V1 through V N consists of certain anticipated radar return samples which were grouped in a 3X3 array. But as an alternative, various other groupings can be used.
  • each unprocessed vector V i could be a 4X5 array of pixels, or a 7X6 array of pixels, etc.
  • the sampled data function f(m,n) is a two-dimensional function.
  • the function by which each unprocessed vector V i is convolved can have any number of dimensions.
  • the symbol f( ) is used.
  • the total number of unprocessed vectors is from sixteen to ten thousand; the total number of samples in each unprocessed vector is from two to one hundred; and the total number of points in said sampled data function is from ten to fifty thousand.
  • the size of the memory M which stores the preconvolved vectors becomes practical with today's technology. For example, suppose the number of unprocessed vectors V1 thru V N is one hundred; each such vector is a 3X3 matrix; f1(m) is a fifty point function; and f2(n) is a fifty point function. Then, the total number of pixels in the preconvolved vectors equals (100)(50+3-1)(50+3-1) or 270,400. Assuming twelve bits per pixel, this equals 3,244,800 bits; and today, a single DRAM chip holds one million bits.
  • the single computer 73 in FIG. 9 can be replaced with multiple computers which operate in parallel. With that arrangement, the offset additions of FIG. 8 could be performed on several overlapping pixels simultaneously. Also, to reduce memory size for the case where f2(n) varies with m, the memory M of FIG. 9 can be double buffered. One buffer would be filled with the preconvolved vectors for the current range of indexes, while the other buffer will be updated with the preconvolved vectors for the next range of indexes.
  • radar return signals were processed to produce visual images.
  • linear signal processing can be performed in any system - radar or nonradar - which previously performed FFT's or inverse FFT's or convolutions in real time.
  • Such systems include, but are not limited to, speech signal processors, seismic signal processors, image signal processors, synthetic aperture sonar processors, infrared image processors, automatic image recognition systems, and spectrum analyzers.

Abstract

Un système électronique de traitement de signaux d'entrée de données échantillonnées comprend une mémoire électronique stockant un ensemble de vecteurs pré-traités V1*f( ), V2*f( ),...VN*f( ) dans lequel f( ) est une fonction de données échantillonnées, * représente un opérateur de convolution, et V1 à VN représentent un ensemble fini de vecteurs non traités N, chacun desquels représente un groupe anticipé d'échantillons de signaux d'entrée. Une fois ces vecteurs pré-traités stockés, on traite des signaux d'entrée en temps réel au moyen (a) d'un circuit (20b) échantillonnant le signal d'entrée; (b) d'un circuit (20d) comprimant la séquence d'échantillons transformée en une séquence plus petite de signaux d'indices correspondant aux indices 1 à N des vecteurs non traités; (c) d'un circuit (20f) recevant la sequénce plus petite de signaux d'indices, et lisant à partir de la mémoire (M) les vecteurs prétraités dont les indices correspondent aux signaux d'indices reçus; et (d) un circuit (20h) additionnant ensemble les vecteurs pré-traités lus tout en maintenant un décalage prédéterminé entre eux à mesure qu'ils sont additionnés.

Claims (11)

  1. Dispositif électronique de traitement de données pour la sommation de données prétraitées décalées à partir d'une mémoire,
       ledit dispositif étant caractérisé par le fait qu'il comprend :
    - un moyen de mémoire électronique (M sur la FIGURE 9) stockant un ensemble de vecteurs prétraités V₁*f(m, ...) à Vn*f(m, ...) où f(m,...) est une fonction de données échantillonnées possédant un nombre quelconque de dimensions m, ..., * est un opérateur de combinaison et V₁ à VN est un ensemble fini de N vecteurs non traités représentant chacun un groupe anticipé d'échantillons du signal d'entrée;
    - un moyen (71, 72 sur la FIGURE 9) pour la génération d'une séquence d'échantillons à partir d'un signal d'entrée devant être traité et pour la compression de ladite séquence d'échantillons en une séquence plus petite de signaux d'index correspondant aux index 1 à N desdits vecteurs non traités;
    - un moyen (73 sur la FIGURE 9; 82 sur la FIGURE 10) pour la réception de ladite séquence plus petite de signaux d'index et pour la lecture à partir dudit moyen de mémoire des vecteurs prétraités dont les index correspondent aux signaux d'index reçus; et
    - un moyen (73, 74 sur la FIGURE 9; 83 sur la FIGURE 10) pour la sommation des vecteurs prétraités lus tout en maintenant un décalage prédéterminé entre eux pendant leur sommation.
  2. Dispositif selon la revendication 1, dans lequel ledit signal d'entrée est un signal de retour d'une impulsion radar (20b sur la FIGURE 1).
  3. Dispositif selon la revendication 1, dans lequel le nombre de vecteurs prétraités (V₁* à Vn* sur la FIGURE 9) s'étend de 16 à 10 000.
  4. Dispositif selon la revendication 1 ou 2, dans lequel ledit décalage prédéterminé (50, 51 sur la FIGURE 8) qui est maintenu entre les vecteurs prétraités est égal à la taille et à la forme d'un vecteur non traité.
  5. Dispositif selon la revendication 1, dans lequel ladite fonction de données échantillonnées f (m, ...) (10c sur la FIGURE 1) s'étend de 10 à 50 000 points de données.
  6. Dispositif selon la revendication 1 ou 4, dans lequel ladite fonction de données échantillonnées est de la forme f₁(m)f₂(n) et dans lequel f₁(m) est combinée par rangée avec chaque vecteur V₁ à Vn tandis que f₂(n) est combiné par colonne avec chaque résultat d'une telle combinaison de f₁(m) (FIGURE 6 et FIGURE 7).
  7. Dispositif selon la revendication 1, dans lequel chacun desdits vecteurs non traités V₁ à Vn (FIGURE 4) se rapproche de 2 échantillons à 100 échantillons dudit signal d'entrée.
  8. Dispositif selon la revendication 1 ou 6 dans lequel ladite fonction de données échantillonnées comprend la réponse d'impulsion échantillonnée d'un filtre de gamme (équation 2 sur la FIGURE 6).
  9. Dispositif selon la revendication 1 ou 8, dans lequel ladite fonction de données échantillonnées comprend la réponse d'impulsion échantillonnée d'un filtre d'azimut (équation 3 sur la FIGURE 7).
  10. Dispositif selon la revendication 1 ou 9, dans lequel ledit moyen de lecture et ledit moyen de sommation (73, 74 sur la FIGURE 9; 82 et 83 sur la FIGURE 10) fonctionnent en séquence sur un vecteur prétraité à la fois.
  11. Dispositif selon la revendication 1 ou 10, dans lequel ledit moyen de lecture et ledit moyen de sommation (73, 74 sur la FIGURE 9; 82 et 83 sur la FIGURE 10) fonctionnent en parallèle sur de multiples vecteurs prétraités à la fois.
EP89904671A 1988-02-17 1989-02-17 Procede et appareil de traitement de signaux de donnees echantillonnees Expired - Lifetime EP0355158B1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US157199 1988-02-17
US07/157,199 US4977604A (en) 1988-02-17 1988-02-17 Method and apparatus for processing sampled data signals by utilizing preconvolved quantized vectors

Publications (2)

Publication Number Publication Date
EP0355158A1 EP0355158A1 (fr) 1990-02-28
EP0355158B1 true EP0355158B1 (fr) 1993-10-27

Family

ID=22562731

Family Applications (1)

Application Number Title Priority Date Filing Date
EP89904671A Expired - Lifetime EP0355158B1 (fr) 1988-02-17 1989-02-17 Procede et appareil de traitement de signaux de donnees echantillonnees

Country Status (6)

Country Link
US (1) US4977604A (fr)
EP (1) EP0355158B1 (fr)
JP (1) JPH03502602A (fr)
KR (1) KR900700894A (fr)
DE (1) DE68910222T2 (fr)
WO (1) WO1989007773A1 (fr)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR2659817B1 (fr) * 1990-03-13 1992-05-29 Thomson Csf Dispositif de reconnaissance de sequences dans un signal multidimensionnel.
ATE274017T1 (de) * 1991-06-19 2004-09-15 Akzo Nobel Nv Harze auf der basis von epihalohydrin mit verringertem halogengehalt
US5710839A (en) * 1994-04-20 1998-01-20 Eastman Kodak Company Method and apparatus for obscuring features of an image
US20020169735A1 (en) * 2001-03-07 2002-11-14 David Kil Automatic mapping from data to preprocessing algorithms
US20020129342A1 (en) * 2001-03-07 2002-09-12 David Kil Data mining apparatus and method with user interface based ground-truth tool and user algorithms
US7307580B2 (en) * 2006-01-17 2007-12-11 Raytheon Company Non-statistical method for compressing and decompressing complex SAR data
KR102403307B1 (ko) 2016-08-19 2022-05-30 소니그룹주식회사 곱합 연산 장치

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4025920A (en) * 1972-09-28 1977-05-24 Westinghouse Electric Corporation Identification of radar systems
US4219811A (en) * 1975-02-07 1980-08-26 Hughes Aircraft Company Synthetic array autofocus system
US4200861A (en) * 1978-09-01 1980-04-29 View Engineering, Inc. Pattern recognition apparatus and method
US4400828A (en) * 1981-03-27 1983-08-23 Bell Telephone Laboratories, Incorporated Word recognizer
AU554285B2 (en) * 1983-09-06 1986-08-14 Toshiba, Kabushiki Kaisha Image signal system using synthetic aperture technique
JPS6055280A (ja) * 1983-09-06 1985-03-30 Toshiba Corp 開口合成映像化装置
JPS6055281A (ja) * 1983-09-06 1985-03-30 Toshiba Corp 開口合成映像化装置
JPS60159989A (ja) * 1984-01-30 1985-08-21 Hitachi Ltd 特徴パタ−ンによるデ−タ検索装置
US4679164A (en) * 1984-12-17 1987-07-07 The United States Of America As Represented By The Secretary Of The Army Digital high speed programmable convolver

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
COLLOQUE INTERNATIONAL SUR LE RADAR, 4-8 December 1978, Paris, Session IX : "Traitement de Signal - Technologies", Part 3, Communication IX.3; P. ANTHOUARD et al. : "Radar à antenne latérale synthétique et systèmes de traitement associés", pages 295-302. *
PROCEEDINGS OF THE 1988 IEEE NATIONAL RADAR CONFERENCE, 20-21 April 1988, Michigan, IEEE, New York, US; C.J.READ et al.: "A computation compression technique for SAR based on vector quanization", pages 123-128 *
PROCEEDINGS OF THE IEEE, vol. 73, no. 11, November 1985, IEEE, New York, US; J. MAKHOUL et al. : "Vector quantization in speech coding", pages 1551-1588 *

Also Published As

Publication number Publication date
DE68910222D1 (de) 1993-12-02
KR900700894A (ko) 1990-08-17
WO1989007773A1 (fr) 1989-08-24
US4977604A (en) 1990-12-11
EP0355158A1 (fr) 1990-02-28
JPH03502602A (ja) 1991-06-13
DE68910222T2 (de) 1994-03-03

Similar Documents

Publication Publication Date Title
US4189748A (en) Video bandwidth reduction system using a two-dimensional transformation, and an adaptive filter with error correction
US3983320A (en) Raster display histogram equalization
US4513440A (en) Hardware median filter
US5168375A (en) Image reconstruction by use of discrete cosine and related transforms
EP0286183B1 (fr) Système de transmission de télévision utilisant le codage par transformée
US5526299A (en) Method and apparatus for encoding and decoding using wavelet-packets
US4132989A (en) Azimuth correlator for real-time synthetic aperture radar image processing
US5959567A (en) Method and apparatus for tracking of organized storms
US4947446A (en) Method and circuit for filtering signals representative of a picture
EP0289401A2 (fr) Démodulateur numérique pour signaux multiplexés par division en fréquence (FDM)
EP0480507B1 (fr) Méthode et appareil pour mesurer les caractéristiques d'un canal de transmission et restaurer un signal par émission d'un signal de test utilisé pour calculer la réponse impulsionnelle du canal
EP1665077A2 (fr) Moteurs de compression de donnees et compresseur large bande en temps reel pour donnees multidimensionnelles
US3952186A (en) Apparatus for the generation of a two-dimensional discrete fourier transform
EP0740269A2 (fr) Mise à échelle d'images comprimées
EP0355158B1 (fr) Procede et appareil de traitement de signaux de donnees echantillonnees
EP0286184B1 (fr) Transformateur d'images linéaire unidimensionnel
EP0227614A2 (fr) Dispositif de traitement de signaux pour radar à ouverture synthétique, en particulier adapté au calcul parallèle
EP0182243A2 (fr) Filtrage de signal d'image
EP3724680A1 (fr) Système et procédé de formation d'images radar à synthèse d'ouverture
US4837619A (en) Scan rate conversion apparatus and method
US5101445A (en) Method and apparatus for filtering digital data by concatenating preprocessed overlapping groups of the data
US4760540A (en) Adaptive interference cancelling system and method
EP0566184A2 (fr) Transformateur d'image ainsi que système de télévision comportant un émetteur et un récepteur munis d'un transformateur
US5349547A (en) Bidimensional FIR filter
US3652980A (en) Method and apparatus for seismic signal pattern discrimination

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): BE DE FR GB NL

17P Request for examination filed

Effective date: 19900208

17Q First examination report despatched

Effective date: 19920421

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): BE DE FR GB NL

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: NL

Effective date: 19931027

Ref country code: BE

Effective date: 19931027

ET Fr: translation filed
REF Corresponds to:

Ref document number: 68910222

Country of ref document: DE

Date of ref document: 19931202

NLV1 Nl: lapsed or annulled due to failure to fulfill the requirements of art. 29p and 29m of the patents act
PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed
REG Reference to a national code

Ref country code: GB

Ref legal event code: IF02

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: GB

Payment date: 20020108

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20020131

Year of fee payment: 14

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20020228

Year of fee payment: 14

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20030217

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20030902

GBPC Gb: european patent ceased through non-payment of renewal fee
PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20031031

REG Reference to a national code

Ref country code: FR

Ref legal event code: ST